In this article, the authors proposed a robust approach to solve the cooperative airborne VSLAM (visual simultaneous localization and mapping) problem based on the development of a full stereo camera observation model. An adaptive SIFT feature extractor followed by stereo vision constraints fast matching were introduced to construct a large map. Robust C-VSLAM (cooperative VSLAM) is implemented based on the NH∞ filter and compared with the single VSLAM. The proposed C-VSLAM solution was validated using simulation and experimental data and good and very promising results were obtained. Future work will include extending this cooperative strategy in a distributed estimation framework for an experimental setup based on a set of cooperative UAVs (unmanned aerial vehicles).
Robust cooperative visual localization with experimental validation for unmanned aerial vehicles
Robuste kooperative visuelle Lokalisierung mit experimenteller Validierung für unbemannte Luftfahrzeuge
2013
19 Seiten, 21 Bilder, 46 Quellen
Article (Journal)
English
Robust cooperative visual localization with experimental validation for unmanned aerial vehicles
SAGE Publications | 2013
|Robust cooperative visual localization with experimental validation for unmanned aerial vehicles
Online Contents | 2013
|Cooperative target localization using heterogeneous unmanned ground and aerial vehicles
Tema Archive | 2010
|Autonomous target detection and localization using cooperative unmanned aerial vehicles
Tema Archive | 2009
|